Ventilation Box design - Nvidia Jetson Orin

3D Enclosure Design for Thermal Optimization of Edge-AI DevicesThis dataset contains 3D design files (Skp.) of custom enclosure prototypes developed to evaluate airflow strategies for embedded GPU systems running edge-AI workloads. Two enclosure versions are included:Version 1 (Undirected Airflow): initial prototype with fan mounted on the side louver.Version 2 (Directed Airflow): optimized prototype with chimney-style airflow path and improved thermal performance.The enclosures were tested under real-world inference workloads to assess their impact on GPU/CPU temperature, inference latency, and throughput. Results show that active airflow reduces CPU and GPU core temperature by up to 36.45% and 38.65% respectively, compared to the baseline, while sustaining a 9.2% faster and more stable inference.These files are provided to support reproducibility and further research in thermal–performance co-design for edge-AI systems.